Despite the dual role for TGF-beta as both tumor suppressor and tumor promoter in carcinogenesis, preclinical data from our lab and others has previously suggested that strategies to antagonize TGF-beta may selectively reduce the undesirable tumor promoting effects of this growth factor while sparing the desirable effects on tumor suppression and normal homeostasis. Based on these promising preclinical results, several different TGF-beta pathway antagonists are in early phase clinical trials for the treatment of advanced cancer. However, given the complex biology of TGF-beta, the successful development of TGF-beta antagonists for cancer therapy will depend on a clear understanding of how these agents work, and the related question of how to select patients who will benefit from this type of treatment. In our previous work, we performed detailed mechanistic analysis of the mode of action of anti-TGF-beta neutralizing antibodies in the widely used 4T1 transplantable mouse model of metastatic breast cancer. Using a panel of 12 transplantable syngeneic mouse models of metastatic breast cancer, with metastatic burden as the primary therapeutic endpoint, we had previously uncovered heterogeneous responses to TGF-beta antagonism, with inhibition of metastasis in some models, and no effect on or stimulation of metastasis in other models. We have applied discovery-based approaches to address molecular and biological mechanisms underlying the heterogeneity of therapeutic response and to generate useful predictive biomarkers. Whole exome DNA sequencing of the tumor cell lines has shown that none of the commonly occurring mutations in breast cancer are correlated with response to therapy. Transcriptomic analysis of untreated primary tumors in the panel segregates tumors by response to anti-TGF-beta therapy, suggesting that the therapeutic response is dictated by prominent, readily identifiable molecular and biological features of the tumor. Tumors from models showing a desirable response to TGF-beta antagonism are characterized by reduced immune function, enhanced angiogenesis, higher tumor cell proliferation and survival, and transcriptomic evidence of TGF-beta pathway activation in the untreated state. Gene expression signatures that are predictive of response to anti-TGF-beta therapy have been generated and are being tested in additional preclinical models. In our search for useful predictive and pharmacodynamic biomarkers of TGF-beta antagonism, we have developed approaches for quantitative and more sophisticated monitoring of TGF-beta pathway activation in tumors, adapting the ProteinSimple SimpleWestern technology for accurate quantitation of Smad activation in tumor samples, and developing a quantitative brightfield proximity ligation assay for detection of the non-canonical mixed Smad signaling complexes that are associated with pathological TGF-beta signaling. We have also begun to address whether it might be possible to improve the therapeutic response to TGF-beta antagonism by selective neutralization of different TGF-beta isoforms. Correlative clinical data and literature evidence suggests that whereas TGF-beta1 is primarily associated with poor outcome, TGF-beta3 may actually oppose TGF-beta1 and be associated with good outcome, providing a rationale for selective neutralization of TGF-beta1 and TGF-beta2, while sparing TGF-beta3. As part of this initiative, we have developed methods for accurate quantitation of TGF-betas in tumor extracts since we have shown that TGF-beta mRNA levels do not correlate well with protein levels for TGF-beta1 and TGF-beta3. We have assessed TGF-beta isoform protein levels across the metastatic tumor panel and find that TGF-beta1:TGF-beta3 ratios vary over a 20-fold range. Therapeutic anti-TGF-beta antibodies with different isoform selectivity will be compared for efficacy in representative models. Through this approach, we hope to generate improved TGF-beta-targeted therapeutics.